Urokinase-type plasminogen activator (uPA) is a multidomain serine protease, having a catalytic "B" chain (amino acids 144-411), and an amino-terminal fragment ("ATF", aa 1-143) consisting of a growth factor-like domain (4-43) and a kringle (aa 47-135). The uPA kringle appears to bind heparin, but not fibrin, lysine, or aminohexanoic acid. The growth factor-like domain bears some similarity to the structure of epidermal growth factor (EGF), and is thus also referred to as an "EGF-like" domain. The single chain pro-uPA is activated by plasmin or other proteases, cleaving the chain into the two chain active forrn, which is linked together by a disulfide bond.
uPA binds to its specific cell surface receptor (uPAR). The binding interaction is apparently mediated by the EGF-like domain (S.A. Rabbani et al., J Biol Chem (1992) 267:14151-56). Cleavage of pro-uPA into active uPA is accelerated when pro-uPA and plasminogen are receptor-bound. Thus, plasmin activates pro-uPA, which in turn activates more plasmin by cleaving plasminogen. This positive feedback cycle is apparently limited to the receptor-based proteolysis on the cell surface, since a large excess of protease inhibitors is found in plasma, including .alpha..sub.2 antiplasmin, and PAI-1.
Plasmin can activate or degrade extracellular proteins such as fibrinogen, fibronectin, and zymogens, particularly of the matrix metalloproteinases. Plasminogen activators thus can regulate extracellular proteolysis, fibrin clot lysis, tissue remodeling, developmental cell migration, inflammation, and metastasis. Accordingly, there is great interest in developing uPA inhibitors and uPA receptor antagonists. E. Appella et al., J Biol Chem (1987) 262:4437-40, determined that receptor binding activity is localized in the EGF-like domain, and that residues 12-32 appear to be critical for binding. The critical domain alone (uPA.sub.12-32) bound uPAR with an affinity of 40 nM (about 100 fold less than intact ATF).
Recent studies have shown that the invasiveness of human tumor cell lines in vitro correlates with surface bound urokinase, and that urokinase production itself is an independent prognostic indicator in human breast cancer (W. Schlechte et al., Cancer Comm (1990) 2:173-79; H. Kobayashi et al., Br J Cancer (1993) 67:537-44; J. A. Foekens et al., Cancer Res (1992) 52:6101-05). It has also been shown in both breast and colon cancer that urokinase is often made by stromal cells (fibroblasts and macrophages), whereas the urokinase receptor is found on tumor cells (C. Pyke et al., Cancer Res (1993) 53:1911-15; C. Pyke et al., Am J Path (1991) 138:1059-67). UPAR has independently been identified as a monocyte activation antigen, Mo3, whose expression is induced in these inflammatory cells upon activation (H.Y. Min et al., J Immunol (1992) 148:3636-42), as well as an activation antigen on human T lymphocytes (A. Nykjxr et al., J Immunol (1994) 152:505-16). Urokinase plasminogen activator "knock-out" mice (in which the uPA gene is inactivated or deleted throughout the body) have been developed, and their macrophages are deficient in extracellular matrix degradation in vitro (P. Carmeliet et al., Fibrinolysis (1993) 7 Suppl. 1:27-28). In addition, these mice show greatly reduced smooth muscle cell migration/proliferation after arterial wounding, suggesting a possible role for uPA/uPAR in post-angioplasty restenosis.
The induction of urokinase and its receptor by agents known to be angiogenic in vivo, such as bFGF, vEGF, and TNF.alpha., suggests a role for cell surface urokinase in angiogenesis (P. Mignatti et al., J Cell Biol (1991) 113:1193-202; L. E. Odekon et al., J Cell Physiol (1992) 150:258-63; M. J. Niedbala et al., Blood (1992) 79:678-87). Although many factors are likely to be angiogenic in pathological conditions, degradation of extracellular matrix by capillary endothelial cells and release of matrix-bound pro-angiogenic factors by cell surface plasmin is likely a common step in these processes (D. Weinstat-Saslo et al., FASEB J (1994) 8:401-07). This is further supported by the observation that several known anti-angiogenic substances reduce uPA expression (S.
Takano et al., Cancer Res (1994) 54:2654-60). In vivo studies have shown that prevention of urokinase-receptor binding, by urokinase antibodies or competition with inactive urokinase mutants, dramatically reduces or eliminates the metastatic potential of human prostate tumor cells in nude mice (C. W. Crowley et al., Proc Natl Acad Sci USA (1993) 90:5021-25; L. Ossowski et al., Cell (1983) 35:611-19; L. Ossowski, J Cell Biol (1988) 107:2437-45). It has recently been shown in both in vitro and syngeneic in vivo models that the protein uPAR antagonists are anti-angiogenic (Mn et al., Cancer Res (1996) 56:2428).
Although a primary role of uPAR is in the focusing of uPA dependent plasminogen activation to the cell surface, it also has other functions. For instance, uPAR is involved in cell adhesion, functioning as a uPA dependent vitronectin receptor (Wei et al., J Biol Chem (1994) 269:32380-88). More recently, it has been shown that uPAR interacts with integrins and is likely involved in cell shape changes and cell migration (Kindzelskii et al., J Immunol (1996) 156:297).
To date, only two small molecules have been described which inhibit the uPA:uPAR interaction (suramin: N. Behrendt et al., J Biol Chem (1993) 268:5985-89; and 8-anilinonaphthalene sulfonic acid: M. Ploug et al., Biochemistry (1994) 33:8991-97). Unfortunately, these compounds are effective only at micromolar concentrations.